What this site is
moonboard.io explains how the intelligence works. moonboard.ai lets you work with it.
Moonboard is a crypto market intelligence platform. It scans several hundred digital assets, combines market and social data, and reduces each asset to a small set of comparable metrics.
The application lives at moonboard.ai. This domain is the documentation layer underneath it: the formulas, the input fields, the processing chain, and — at least as prominently — the conditions under which each number stops being reliable.
Why the two are separate
A product page and a methodology page are written for different readers with different questions, and putting both on one domain serves neither well.
| moonboard.ai | moonboard.io | |
|---|---|---|
| Purpose | Interactive analysis | Documentation and research |
| Question | "How do I use this?" | "How does this work?" |
| Content | Dashboards, charts, rankings, news | Formulas, field definitions, limitations |
| Data | Live, continuously recomputed | None — this site holds no market data |
| Reader | Anyone analysing assets | Analysts, researchers, developers |
Nothing here duplicates the application's help pages. Where those explain how to use a feature, this site explains what the feature computes and where the computation has limits.
Editorial standards
These are the rules this documentation is written to. They are stated publicly so that failures to meet them can be pointed out.
- Document the implementation, not the intention. Every formula here was checked against the code that runs in production. Where the implementation differs from how a metric is commonly described — as with SDE, which compares sentiment against volatility rather than price direction — this site describes what the code does.
- Limitations are content, not disclaimers. Every methodology page has a limitations section, and it names things that make the metric look worse. A metric whose weaknesses are hidden cannot be used well.
- No invented results. No fabricated studies, no citations to papers that were not read, no statistics presented as measured that were not measured. Research notes is currently empty for exactly this reason.
- No claimed sources that are not used.Data sources documents source categories and refresh behaviour, and says plainly why individual vendors are not named.
- Nothing is described as available before it is.The Data API is marked in development because it is, rather than being documented as though it were finished.
- No page exists only to rank. It would be straightforward to generate hundreds of thin pages for search traffic. The value of this domain has to come from the documentation being genuinely useful, or it has none.
What Moonboard is not
Moonboard produces analytical tools. It does not produce investment advice, and the distinction is not a legal formality:
- No metric is a buy or sell signal. The scores describe observed market and social conditions. A high score means an asset currently exhibits a described pattern — not that it will rise.
- No metric forecasts anything. Every input is a trailing measurement. There is no forward-looking term in any formula on the platform.
- Model output is not fact. Sentiment is a classifier's judgement about text. Liquidation visualisations are reconstructions under assumed leverage. Both are presented as estimates here, because that is what they are.
Who operates this
Moonboard is built and operated by NEOB Technology GmbH, a technology company based in Vienna, Austria, trading as NEOB. The same team builds and runs both moonboard.ai and this site.
Corrections are welcome and useful. If something here is wrong, unclear, or no longer matches what the platform computes, write to office@neob.ai.
Related
- MethodologyStart here to understand the metrics
- ResearchWhere measurement gets hard
- Data categoriesWhat is collected
- DevelopersProgrammatic access status